According to the development of 5+2 industry innovation plan in Taiwan, the manufacturing industry is gradually transforming from the traditional machinery industry to automation and intelligence. Demands for customized products and rare commodities have increased, and the manufacturing line has also been transformed into low volume and high variability mode. The manufacturing efficiency and high processing quality requirements are gradually increasing. Traditional manufacturing relying on manual prototyping, monitoring and debugging is inefficient. As such, in this article, we propose an autonomous decision making system based on experimental determination for the milling process to optimize machining parameters. By analyzing the physical characteristics of the tool and the work-piece and collecting on-line cutting data, the system will automatically output the fastest process parameters based on the stable cutting quality. The method is used to assist operators to quickly obtain the most stable and fastest process parameters. Therefore, the result can increase the efficiency of manufacturing and the mass production.